Lungwort Safety and What Is Genuinely Unknown

A safety page for lungwort has to do three things that a normal safety page does not. It has to keep two completely unrelated organisms apart, because both are sold under this name and their hazards have nothing in common. It has to describe a family-level alkaloid question honestly — neither inflating it into a scare nor waving it away — while refusing to print numbers that the published measurements cannot support. And it has to say clearly what is not wrong with lungwort, because a reader arriving from the comfrey, coltsfoot or horsetail pages will bring specific worries with them, and several of them do not apply.

The short version, before the detail. Pulmonaria officinalis looks like a mild herb, and the honest verdict on it is caution plus an unresolved question rather than either reassurance or alarm. Lobaria pulmonaria, the lichen, has a different profile entirely, no toxicology at all, and a specific contamination problem that follows from the way lichens are built. Neither has ever been the subject of a safety study in a human being.

Table of Contents

  1. Two Organisms, Two Safety Profiles
  2. Pyrrolizidine Alkaloids: What They Are and How They Injure
  3. What Has Actually Been Measured in Pulmonaria
  4. Lungwort Is Not Comfrey, and That Cuts Both Ways
  5. The Contamination Nobody Expects: Co-Harvested Weeds
  6. Allantoin: A Real Compound With the Wrong Route
  7. Silica, Tannins and Two Smaller Claims
  8. The Lichen's Own Safety Profile
  9. Why a Wild Lichen Is an Uncontrolled Exposure
  10. Who Should Not Use It At All
  11. Hazards Lungwort Does Not Have
  12. The Unknowns, as Numbered Findings
  13. Practical Guidance If You Use It Anyway
  14. Key Research Papers
  15. Connections

Two Organisms, Two Safety Profiles

As the identity page establishes, “lungwort” names both a borage-family flowering plant and a lichen, and their chemistries do not share a single compound class. It follows that they do not share a safety profile either, and merging the two — in either direction — produces a wrong answer twice.

Every claim below is labelled with the organism it belongs to. If a sentence does not name the lichen, it is about the plant.

Pyrrolizidine Alkaloids: What They Are and How They Injure

Pyrrolizidine alkaloids are esters built on a bicyclic amine called a necine base, made by plants across several unrelated families — Boraginaceae, Asteraceae tribes including Senecio and Eupatorium, and Fabaceae — apparently as a defence against insects. They are among the best-characterised plant toxins in human medicine, and the mechanism is worth understanding because it explains every one of the practical cautions further down.

Four points carry the argument.

  1. Only some of them matter. Toxicity belongs to the 1,2-unsaturated alkaloids — those built on retronecine, heliotridine or otonecine bases — and to their N-oxides. Saturated types are not considered a hazard. So “contains pyrrolizidine alkaloids” is an incomplete statement until the type is named.
  2. The liver does the damage to itself. The alkaloids are not directly toxic. Hepatic cytochrome P450 enzymes, prominently CYP3A4, convert them into reactive pyrrolic esters, which alkylate proteins and DNA in the cells that made them. The organ that metabolises the compound is the organ that is injured.
  3. The characteristic injury is vascular. The reactive metabolites damage the endothelial lining of the small hepatic veins and sinusoids, producing hepatic sinusoidal obstruction syndrome — the condition formerly called veno-occlusive disease. Acute high exposure gives abdominal pain, an enlarged tender liver, fluid in the abdomen and jaundice. Chronic low-level exposure can instead produce insidious fibrosis and eventually cirrhosis, with no distinctive early signs at all.
  4. Some are genotoxic. Several 1,2-unsaturated alkaloids are mutagenic in test systems and carcinogenic in rodents, and individual members have been evaluated internationally as possible human carcinogens. This is why regulators treat them as compounds with no clearly safe intake rather than compounds with a tolerable daily dose.

Potency varies enormously by structure, and this is where lungwort's position is genuinely different from comfrey's. The macrocyclic diesters of the Senecio type are generally the most potent. The open-chain diesters — echimidine and lasiocarpine, found in comfrey, borage and heliotrope — come next. The lycopsamine-type monoesters are generally regarded as the least potent subclass. They are still 1,2-unsaturated, still counted in regulatory risk assessments, and not to be dismissed — but they are not the compounds behind the classic poisoning outbreaks. The alkaloids reported in Pulmonaria are of the lycopsamine type.

Regulators have taken the class seriously. European food-safety assessments of pyrrolizidine alkaloids in honey, tea, herbal infusions and food supplements have used a margin-of-exposure approach anchored on rodent carcinogenicity data and concluded that habitual high consumption of contaminated herbal products is a possible concern; maximum levels for pyrrolizidine alkaloids in herbal infusions and supplements subsequently entered EU food law in the early 2020s. We are deliberately not quoting a threshold figure — see the next section for why. The current documents are on the European Food Safety Authority site, and the National Institutes of Health LiverTox resource carries a clinical chapter on the class.

What Has Actually Been Measured in Pulmonaria

Here is the whole of it, which is less than most sources imply.

We are not going to print a content figure, and this is why. The direct measurement on P. officinalis is from 1984 by older methodology; the detailed part-by-part gradient is from a different species; nothing has been published on commercial dried material, on seasonal variation, or on how much transfers into an infusion. Sources that quote a confident microgram-per-gram figure for lungwort are, as far as we can establish, extrapolating. A number invented here would be worse than no number, so there is none. What we can say is the shape of the finding: alkaloids of a lower-potency subclass have been reported in the genus, with the highest concentrations in the underground parts and low levels in the leaf.

And the honest verdict, stated as a verdict rather than left implicit: family membership raises a real question about Pulmonaria officinalis, and is not proof about it. The position is caution plus an unresolved question. Anyone who tells you the question is settled — in either direction — is going beyond the data.

Lungwort Is Not Comfrey, and That Cuts Both Ways

Comfrey is the borage-family plant whose alkaloid problem is thoroughly documented: higher total content, including the more potent open-chain diester echimidine, and — decisively — published human cases of hepatic sinusoidal obstruction syndrome following ingestion. That evidence is why comfrey is restricted or prohibited for internal use in several countries and why this site's comfrey page is titled the way it is.

Direction one: do not transfer comfrey's case reports to lungwort. Different species, a different and lower-potency alkaloid subclass, and by all indications a much lower content. Attaching comfrey's liver injuries to lungwort would be exactly the borrowed-evidence error this site flags elsewhere, and it would be manufacturing a hazard rather than reporting one. Naming what a herb is not guilty of is part of accuracy.

Direction two: do not read the absence of lungwort case reports as a clean record. There is no published human case of lungwort-associated liver injury that we could find. That is a weak reassurance for four specific reasons:

  1. Exposure is tiny. Lungwort is a minor herb. A hazard cannot generate case reports from a population that barely exists.
  2. The injury is anonymous. Chronic low-dose pyrrolizidine injury presents as fatigue, vague abdominal discomfort, then fibrosis. There is no rash, no timing signature, nothing that makes a clinician ask about tea.
  3. The latency is long. Exposure and diagnosis can be separated by years, by which time the herbal history is gone.
  4. Nobody is looking. There is no surveillance programme, no product testing requirement specific to this herb, and no poison-centre series.

So the argument for caution does not rest on a case report. It rests on asymmetry: the potential downside is liver injury that is silent until it is advanced, and the potential upside is a soothing tea whose respiratory benefit, as the respiratory page documents, has never been demonstrated. When one side of a bet is unquantified harm and the other is unquantified nothing, short and occasional use is the rational position even with no documented injury at all.

The Contamination Nobody Expects: Co-Harvested Weeds

There is a twist in the pyrrolizidine story that applies to lungwort regardless of what Pulmonaria itself contains, and it is the single most practically useful thing on this page.

When European authorities began surveying herbal teas for pyrrolizidine alkaloids with modern methods, they found them widely — including in teas made from plants that do not produce them at all: chamomile, peppermint, rooibos, black and green tea. The alkaloids came from other plants harvested along with the crop. Ragwort and groundsel (Senecio), viper's bugloss (Echium), heliotrope and hound's tongue grow as field weeds, are mechanically harvested with the herb, and carry alkaloids of the more potent classes.

Two consequences follow, and they point in opposite directions from what most readers would guess:

This is the same structure as the argument on the site's coltsfoot regulation page: the standard exculpation — the harm came from something else in the bag — clears the named plant and simultaneously indicts unverified loose-leaf material. If you cannot establish what is in the bag, the hazard is worse than the labelled species' hazard, not better.

Allantoin: A Real Compound With the Wrong Route

Allantoin is one of lungwort's genuinely interesting constituents, and it is routinely over-claimed, so it deserves careful handling.

What it is. A small, water-soluble heterocyclic compound, reported in Pulmonaria and well known from comfrey. It is a recognised skin-conditioning and keratolytic agent, appears in over-the-counter skin-protectant formulations, and has animal and formulation-level work behind its use in wound care and dermatitis.

The route problem. All of that evidence is topical. Allantoin's proposed actions — softening keratin, promoting epithelial turnover, soothing irritated skin — are surface effects on skin that has the compound applied to it. A different route is a different intervention. Drinking a lungwort infusion does not put allantoin on your bronchi, your throat lining in any therapeutic sense, or anywhere the topical literature has studied. So allantoin supports lungwort's traditional external use on cuts, grazes and haemorrhoids, and contributes nothing to the respiratory claim. It is quietly presented as if it did, fairly often.

The point almost nobody makes. Allantoin is not exotic to human physiology at all — it is an endogenous compound. Humans lack a functioning uricase, so we do not convert uric acid to allantoin enzymatically, but urate is oxidised to allantoin non-enzymatically by reactive oxygen species, and plasma and urinary allantoin are used in research as biomarkers of oxidative stress. You are producing and excreting it right now.

That fact cuts both ways, honestly reported:

Silica, Tannins and Two Smaller Claims

Silicic acid. Lungwort is described in the herbal literature as notably rich in soluble silica, and that appears to be genuine. The claim built on top of it — that swallowed silica “strengthens lung tissue” — is signature reasoning in chemical dress, as the doctrine page discusses. There is no pharmacokinetic work showing preferential delivery of dietary silicon to pulmonary connective tissue, and no trial of any kind. Treat it as a compositional fact with no established consequence.

Tannins. Real, and responsible for the astringent half of lungwort's contradictory traditional reputation. One consequence is worth knowing because it is well established for tannin-rich beverages generally: polyphenols inhibit the absorption of non-haem iron from a meal, measurably so for tea and coffee drunk with food. For an occasional cup of lungwort during a cough this is negligible. For anyone who is iron-deficient and drinking tannin-rich infusions habitually with meals, it belongs in the ledger. The general fix is the same as for tea: drink it between meals rather than with them.

We are also not going to reproduce per-hundred-gram composition tables for this plant. A 2 g infusion of dried leaf is a small fraction of any per-100 g figure, and only part of that extracts into water, so such figures are ceilings generous to the herb rather than descriptions of what you drink. Vitamin C is the clearest example: it is listed as a constituent, it is heat-labile, and a hot infusion is not a meaningful vitamin C source.

The Lichen's Own Safety Profile

Now the lichen, kept strictly separate.

What it contains. Lobaria pulmonaria's reported secondary metabolites centre on the stictic-acid group of depsidones — stictic, constictic and norstictic acids and relatives — alongside depside-type compounds such as tenuiorin and methyl orsellinate. No pyrrolizidine alkaloids; lichens do not make them. No mucilage of the marshmallow type either, so the demulcent rationale available to the plant is unavailable here.

A borrowed hazard we are refusing to borrow. The obvious move at this point would be to reach for usnic acid, which has a documented and serious liver problem: acute hepatitis and cases of fulminant liver failure were reported in the early 2000s in people taking a usnic-acid-containing weight-loss supplement, published in Annals of Internal Medicine, the American Journal of Gastroenterology and Mayo Clinic Proceedings. It is a real and well-attested hepatotoxin at supplement doses.

But usnic acid is characteristic of other lichen genera — Usnea, Cladonia, Alectoria, Flavoparmelia, Ramalina — and we could not find it reported as a principal constituent of Lobaria pulmonaria, whose chemistry is the stictic-acid series. Attaching those liver-failure cases to tree lungwort would therefore be compound substitution: citing a molecule's toxicity for an organism that is not documented to contain it. That is the same error this site flags when rosmarinic acid effects are credited to whichever plant is being sold. We are not going to commit it in the cautionary direction either, because manufacturing a hazard is as much a failure of accuracy as concealing one.

What the usnic acid story does establish is a general lesson that does apply: lichen secondary metabolites are pharmacologically potent compounds, and one of them caused acute liver failure in otherwise healthy people at supplement doses. “It is only a lichen” is not a safety argument. Against that background, the complete absence of toxicology for Lobaria pulmonaria — no acute study, no repeat-dose study, no human data — is an open gap, not a clean bill of health.

Contact sensitisation is documented for lichens as a group. Lichen substances, usnic acid and atranorin among them, are established contact allergens; the resulting dermatitis has a folk name in forestry — woodcutter's eczema — and oak moss extract is a well-known fragrance allergen for the same reason. Anyone handling quantities of wild lichen should know that repeated skin contact can sensitise.

Why a Wild Lichen Is an Uncontrolled Exposure

This is the concrete hazard of tree lungwort, and it follows from anatomy rather than chemistry.

A lichen has no roots, no vascular system and no waxy cuticle. It takes up water directly across its entire surface, from rain, mist, fog and dust, together with everything dissolved or suspended in it. There is no selective root barrier and no excretory route. That is precisely why lichens are used across the world as biomonitors of air quality: they integrate airborne deposition over years, and their tissue concentrations of metals track the local atmosphere.

The same property makes wild-harvested lichen an uncontrolled exposure:

So “wildcrafted tree lungwort” means: an organism with no toxicology, at an unknown dose, carrying an unknown load of whatever the site's air has been depositing since before it was collected. Add the conservation problem set out on the identity page — this is a slow-growing, old-growth-dependent species on numerous European red lists — and there is no version of this that we can recommend.

Who Should Not Use It At All

These apply to the plant. For the lichen the answer is simpler: there is no basis for internal use by anyone.

One interaction question deserves naming precisely because it is unstudied. Bioactivation of these alkaloids is cytochrome-P450-mediated, CYP3A4 prominent among the enzymes. In principle, anything that induces or inhibits CYP3A4 could shift the balance between activation and clearance. In practice, no interaction study of lungwort exists at all, so this is a mechanistically sensible concern with zero data behind it — which is a reason for caution in the medicated, not a reason to claim a known interaction.

Hazards Lungwort Does Not Have

A reader who arrives here from other pages on this site may be carrying worries that do not belong to this plant. Clearing them is part of getting the safety picture right, so each of the following was checked and comes back negative.

The Unknowns, as Numbered Findings

Written as results rather than gaps, so that “we do not know” is a stated finding instead of something the reader has to infer.

  1. No modern alkaloid analysis of commercial lungwort. No published measurement of total pyrrolizidine alkaloids and N-oxides in dried P. officinalis sold as tea, by current LC-MS/MS panels.
  2. No variation data for the right species. The part-by-part gradient (root high, leaf trace) is from P. obscura. Seasonal, geographic and cultivar variation in P. officinalis is unmeasured — notable given that the plant's phenolic profile is documented to shift between spring and autumn.
  3. No infusion-transfer study. Nobody has measured how much alkaloid moves from dried lungwort leaf into hot water. These compounds and their N-oxides are water-soluble, and substantial transfer has been demonstrated for other herbal teas, so the expectation is meaningful transfer — but that is an expectation, not a measurement for this herb.
  4. No animal toxicology of the leaf. No acute study we would quote, no sub-chronic study, no no-observed-adverse-effect level, no reproductive or developmental toxicology.
  5. No human pharmacokinetics for anything in it. Not for the alkaloids, not for rosmarinic acid, not for allantoin, not for silicon.
  6. No dose-finding and no data-derived duration limit. The “short-term use” advice on this site and elsewhere is inferred from the compound class, not from a study of this plant.
  7. No interaction studies of any kind, including the mechanistically obvious CYP3A4 question. Read “no known interactions” as nobody has looked.
  8. No paediatric, pregnancy or lactation data. The avoidance advice above is entirely class-based.
  9. No surveillance. No poison-centre series, no adverse-event signal analysis, no recall history specific to lungwort. Absence of reports here reflects absent monitoring, not demonstrated safety.
  10. No product identity survey. Nobody has published an analysis of which organism is actually in bags labelled “lungwort.”
  11. For Lobaria pulmonaria, no toxicology whatsoever — and no contaminant survey of commercial material either, despite the accumulation problem described above.
  12. No meaningful definition of an “alkaloid-free” lungwort product, since no specific limit or testing standard has been set for this herb.

Twelve findings. Note that they are all statements of absent evidence, not negative evidence — nothing on the list was tested and failed. Both are unflattering, but they are different epistemic states and conflating them would be sloppy in both directions.

Practical Guidance If You Use It Anyway

This site is educational, not prescriptive, and readers make their own decisions. If lungwort appeals to you, the following reduces the unknowns as far as they can be reduced.

  1. Insist on a binomial. “Lungwort” alone identifies nothing. If the label does not say Pulmonaria officinalis, you do not know whether you have bought a plant or a lichen.
  2. Leaf and flowering top only, never root. The alkaloids are reported to concentrate in the underground parts.
  3. Prefer a tested supplier over a wildcrafted one. Because of the co-harvested-weed problem, testing of the finished product for total pyrrolizidine alkaloids is more informative than any claim about the species.
  4. Water infusion, if the point is soothing. A tincture cannot deliver the demulcent mechanism, since alcohol precipitates mucilage.
  5. Short courses, not a habit. A few days during a cough, not a daily tonic for weeks. Duration is the variable the class hazard responds to.
  6. Do not stack Boraginaceae. Comfrey, borage herb and lungwort draw on the same total exposure.
  7. Observe the avoid list above without exception — pregnancy, breastfeeding, children, liver disease.
  8. Consider whether you want the mechanism or the history. If you want a soothing throat drink with characterised material behind it, marshmallow root, ribwort plantain, mullein and honey deliver the same mechanism with better specifications and no alkaloid question. Lungwort's real attraction is that it is a beautiful plant with a five-century story attached, and that is a legitimate reason to be interested in it — just not a pharmacological one.

When to stop and see a clinician instead. For the chest: breathlessness, chest pain, fever, coughing up blood, unexplained weight loss, or a cough lasting beyond about three weeks. For the liver, and this matters specifically because of the alkaloid class: yellowing of the eyes or skin, dark urine, pale stools, persistent right-upper-abdominal discomfort, abdominal swelling, or unexplained fatigue. Sinusoidal obstruction syndrome has no distinctive early presentation, so if you have been taking a Boraginaceae herb regularly, say so unprompted — nobody will think to ask.

Back to Table of Contents

Key Research Papers

Each entry names the organism, compound or subject actually studied. Links are PubMed topic and title searches rather than fixed record numbers.

  1. [P. officinalis and Borago officinalis] Lüthy and colleagues, “Pyrrolizidine alkaloids in medicinal plants of Boraginaceae: Borago officinalis and Pulmonaria officinalis,” Pharmaceutica Acta Helvetiae, 1984 — PubMed. The only direct measurement we could find on the species actually sold in Europe; four decades old and by pre-LC-MS/MS methodology.
  2. [P. obscura — a different species] Haberer and colleagues, “Pyrrolizidine alkaloids in Pulmonaria obscura,” Planta Medica, 2002 — PubMed. Source of the “lycopsamine-type, mostly in the roots” statement that is routinely quoted as though it were about P. officinalis.
  3. [Boraginaceae, family level] Surveys of pyrrolizidine alkaloid structures and distribution across the borage family — PubMed topic search. Some of this literature sits in journals with patchy PubMed coverage.
  4. [Compound class, toxicology] Moreira and colleagues, “Pyrrolizidine Alkaloids: Chemistry, Pharmacology, Toxicology and Food Safety,” International Journal of Molecular Sciences, 2018 — PubMed. Good entry point for the 1,2-unsaturated distinction and the structural potency ranking.
  5. [Compound class, mechanism] Yan and colleagues, “Identification of Toxic Pyrrolizidine Alkaloids and Their Common Hepatotoxicity Mechanism,” International Journal of Molecular Sciences, 2016 — PubMed. The cytochrome-P450 bioactivation step.
  6. [Boraginaceae, exposure assessment] Zan and colleagues, “Pyrrolizidine alkaloids and health risk of three Boraginaceae used in traditional medicine,” Frontiers in Pharmacology, 2023 — PubMed.
  7. [Compound class, balanced review] Jayawickreme and colleagues, “Pyrrolizidine Alkaloids — Pros and Cons for Pharmaceutical and Medical Applications,” International Journal of Molecular Sciences, 2023 — PubMed.
  8. [Comfrey and related, human injury] Case reports of hepatic veno-occlusive disease following ingestion of pyrrolizidine-containing herbal products, including the comfrey literature from the 1980s onward — PubMed topic search. Labelled as comfrey and relatives, not lungwort.
  9. [Herbal teas, contamination] Surveys of pyrrolizidine alkaloids in herbal teas and infusions, and the co-harvested-weed source — PubMed topic search. The reason product testing beats species reassurance.
  10. [Allantoin, topical] Wound-healing and skin-formulation work on allantoin, including the comparative studies against comfrey root extract — PubMed topic search. Topical route throughout.
  11. [Allantoin, endogenous] Allantoin as a non-enzymatic oxidation product of urate and its use as an oxidative-stress biomarker — PubMed topic search.
  12. [Usnic acid — NOT documented in Lobaria pulmonaria] “Severe hepatotoxicity associated with the dietary supplement LipoKinetix,” Annals of Internal Medicine, 2002; “Fulminant liver failure due to usnic acid for weight loss,” American Journal of Gastroenterology, 2004; and the Mayo Clinic Proceedings report of 2006 — PubMed topic search. Cited as evidence that lichen metabolites can be potently hepatotoxic, not as evidence about tree lungwort, which is not documented to contain usnic acid.
  13. [Lobaria pulmonaria, chemistry] Stictic-acid-group depsidones and related lichen substances in tree lungwort — PubMed topic search.
  14. [Lichens, contact allergy] Lichen-substance contact dermatitis, including usnic acid and atranorin sensitisation and oak moss as a fragrance allergen — PubMed topic search.
  15. [Lichens, accumulation] Lichens as biomonitors of airborne metals, and radiocaesium accumulation in lichens after Chernobyl — PubMed topic search. The basis for treating wild-harvested lichen as an uncontrolled exposure.
  16. [Polyphenols and iron] Hurrell and colleagues on inhibition of non-haem iron absorption by polyphenol-containing beverages, British Journal of Nutrition, 1999 — PubMed.
  17. [Lobaria pulmonaria, toxicology] A direct search for toxicology or safety data on tree lungwort — PubMed topic search — returns nothing usable. That empty result is finding eleven above.

Back to Table of Contents

Connections

Back to Table of Contents